Monitoring solution for two linux servers
Posted by neilyoung57@reddit | linuxadmin | View on Reddit | 27 comments
Hey,
I'm looking for a monitoring solution for two ubuntu servers. Seems to me there is a lot of different solution and I'm getting a bit lost. I'm looking to monitor things such as basic hardware usage, users logs and commands, open ports, security...
We use Entra ID a lot. I wonder if it's worth monitoring those servers with Azure Arc & Azure Monitor for simplicity sakes. Seems rather cheap for two servers. We also already use Defender for all our endpoints (except those servers).
What do you guys use for monitoring ? Can Azure and Defender works well with Linux servers ?
Oliver36832@reddit
I use PRTG, it allows monitoring CPU/RAM/disk status, open ports and syslog collection on two Ubuntu servers from a single console.
Because it retrieves data using standard protocols (SNMP/SSH/ICMP/HTTP), installation is straightforward it's more than adequate for small environments.
spuyet@reddit
https://fivenines.io/ : 5 first monitored servers are free and you can monitor up to 30 servers for 5$ a month
ivansalloum@reddit
I spent a week creating a guide called "Linux Server Resource Monitoring Made Easy". In it, I cover key areas like CPU, memory, storage, and disk I/O. I also go beyond basic monitoring, explaining concepts like load average, process states, memory metrics (e.g., virtual vs. resident memory), context switching, I/O wait, tmpfs filesystems, and how to monitor them. I also explain how to use the du command to analyze directories and identify large files consuming space.
Additionally, I shared an experience where I discovered that a slow disk was causing high I/O wait, which significantly impacted performance.
I hope this guide will help you understand resource monitoring better and give you a solid starting point.
Link: https://ivansalloum.com/linux-server-resource-monitoring-made-easy/
crreativee@reddit
Applications Manager by ManageEngine might be something you're looking for.
leoniscsem@reddit
If you want a convenient and very simple solution, go for Cockpit: https://cockpit-project.org/
And of course, using the fabulous ELK stack will provide you with a lot of joy.
spantosh@reddit
you can try Prometheus + Grafana
0bel1sk@reddit
lgtm
Vuiz@reddit
Loki for logs and Grafana Alloy as agent (instead of node exporters)
Bubbadogee@reddit
+1, great solution
ReactionOk8189@reddit
this is the way
cvilsmeier@reddit
You'll see Prometheus + Grafana mentioned a lot. But if you want a simpler solution you can try https://monibot.io
XrT17@reddit
Zabbix
symcbean@reddit
> Can Azure and Defender works well with Linux servers ?
Having run Ubuntu servers on Azure, I must say that the quality of the azure agent software left a LOT to be desired. That was a couple of years ago....maybe they've got better. OTOH having since moved to a AWS shop, I'm struggling to imagine why anyone would consider Azure for hosting even if the costs were the same.
There's been mention of Prometheus & Grafana. Probably some will mention Loki if its not already here. These are good for collecting analytics - but getting the data out of their databases and into a modelling tool can be a PITA. And they're not very good at managing alerting/escalation/automating actions/discovery.
> basic hardware usage
CPU? Memory? Disk space?...you'd be hard pressed to find something describing itself as a monitoring solution which doesn't handle those.
> users logs and commands,
Whoa - this is a VERY different ball game. Its already available on your boxes - but if you want analytics you will not this from an off the shelf standard monitoring tool - you're talking about a SIEM here.
Since you seem to be using the term "server" in the (inaccurate) vernacular sense to refer to a host, I think it's a bit odd that you made not mention of monitoring the applications on these hosts. If that is a consideration (it *REALLY* should be) then you might want to have a look at Zabbix, Icinga or Check_MK
AxisNL@reddit
Checkmk is great, and you can also deploy it natively on Linux, or as a docker container, or as an appliance in a vm, where no Linux knowledge is needed. And you can also use it to monitor any other machines that you might have.
andyh200@reddit
Have a look at Uptime Kuma, for simple monitoring https://uptime.kuma.pet/
Or for something more in depth try Zabbix https://www.zabbix.com/
derprondo@reddit
Uptime Kuma is the GOAT.
Biervampir85@reddit
CheckMK is another option
jaymef@reddit
netdata for ease
TommyTwoPuds@reddit
I've used Observium before - really easy setup and just havenit discover your network. https://www.observium.org/
cwalls6464@reddit
We use wazuh where im at. It does require agent install though but it will work for everything you listed.
skinney6@reddit
Collectd or Telegraf collection agent running on the servers and shipping to InfluxDB then visualize with Grafana is an option.
Busy-Insurance5822@reddit
Use python scripting its cheaper
knobbysideup@reddit
Naemon for active up/down checks. Graylog for log streaming and analysis.
Formus@reddit
Zabbix + Nagios as open software recommenden solution. For paid option you also have geneos which is a bit nicer on the interface. Or another paid combo could be grafana + dynatrace
Imbrex@reddit
If you are ever considering expanding or monitoring more devices zabbix is my recommendation.
qordita@reddit
Feels like you're looking for logwatch and not an actual monitoring platform.
S0A77@reddit
Elastic